Lower extremity osteoarthritis (OA) increases the risk of falling by up to 2.5 times compared to healthy older adults. However, the mechanisms underlying this increased risk are unknown and there are no strategies specific to people with lower extremity OA to reduce this risk. The long-term goal of the research is to characterize the causes of falls by people with lower extremity OA and to develop strategies that minimize the number of preventable falls in this population. Many symptoms and characteristics of lower extremity OA may interfere with the ability to perform a compensatory stepping response (CSR) in response to large postural disturbances. Thus, the objectives of the present application, which will focus on trip-related falls, are to identify the risk factors that may be responsible fo the increased fall risk of people with lower extremity OA, specifically knee OA, and to establish i people with knee OA will benefit from an intervention that specifically targets the ability to perform CSRs. The central hypothesis is that knee OA impairs the ability to successfully perform a CSR to avoid a trip-related fall, and that the CSR can be improved by task-specific training. The underlying rationale of the proposed research is that identifying modifiable risk factors associated with poor CSR performance will allow clinicians to identify OA patients who may benefit from task-specific CSR training. The central hypothesis will be tested by pursuing three specific aims: (1) Establish the relationships between postural sway measures and trip-related fall outcomes following large postural disturbances for people with knee OA; (2) Quantify the extent to which characteristics and symptoms of knee OA affect CRS performance; and (3) Establish the trainability of the CSR for people with knee OA. For the first aim, measures of static and dynamic postural sway will be collected in subjects with knee OA. These measures will be related to their performance (success or failure) following a laboratory-induced trip and their performance following a treadmill-delivered trip-surrogate. For the second aim, symptoms and characteristics associated with OA (e.g. pain, function, strength) will be assessed in OA and control subjects. Analyses will determine if CSR performance becomes increasingly impaired as negative characteristics associated with knee OA increase in prominence. Finally, in the third aim, subjects with knee OA will respond to a series of treadmill-delivered trip-surrogates that require at least one forward directed step to restore dynamic stability. Improvements in CSR performance-related variables during the series of trials will be assessed to determine the extent to which the CSR is trainable for people with knee OA. The proposed research is significant because it is expected to advance the understanding of fall risk factors in a population with a significantly increased fall risk compared to older adults who do not have OA. The innovation of the proposed research lies in its potential to allow clinical identification of people with knee OA who may benefit from fall prevention interventions using task-specific training.